Statistical and deep learning model for volatility forecasting of Bitcoin in real time in 6G network

In recent years, the cryptocurrency market has gained momentum with Bitcoin emerging as the leading digital asset amidst growing global interest. Bitcoin prices have shown high volatility, which includes a significant level of risk that requires an efficient and robust forecasting model to increase investor trust. Deep learning models such as recurrent neural network (RNN) and long-short term memory (LSTM) are efficient in performing volatility prediction for stock price data. As LSTM shows a long memory effect, it performs better than other time series models. The chapter explores some popular methods like Generalized Autoregressive Conditional Heteroskedasticity (GARCH), Exponential Generalized Autoregressive Conditional Heteroskedasticity (EGARCH), Long Short-Term Memory (LSTM), recurrent neural network (RNN), and Autoregressive Integrated Moving Average (ARIMA) for forecasting volatile data like Bitcoin. A comparison between these models and their effectiveness is done in time series forecasting of such volatile data as Bitcoin. Evaluation metrics like mean square error (MSE), mean absolute error (MAE), mean error (ME), root mean square error (RMSE), Akaike information criterion (AIC) are utilised to measure the performance of forecasting. The chapter also explains the limitations of various time series forecasting models in real-time data streams from Bitcoin markets and time series modelling for forecasting future prices. In addition, the chapter discusses the benefits provided by 6G networks in collecting real-time data from the Bitcoin market.